We introduce APEX-seq, a method for RNA sequencing based on spatial proximity to the peroxidase enzyme APEX2. APEX-seq in nine distinct subcellular locales produced a nanometer-resolution spatial map of the human transcriptome, revealing extensive and exquisite patterns of localization for diverse RNA classes and transcript isoforms. We uncover a radial organization of the nuclear transcriptome, which is gated at the inner surface of the nuclear pore for cytoplasmic export of processed transcripts. We identify two distinct pathways of messenger RNA localization to mitochondria, each associated with specific sets of transcripts for building complementary macromolecular machines within the organelle. APEX-seq should be widely applicable to many systems, enabling comprehensive investigations of the spatial transcriptome.(ERM), which we previously attempted but failed to map using APEX2-ERM and APEX-RIP (Kaewsapsak et al., 2017). When we performed a comparison of APEX-seq and APEX-RIP using HEK 239T cells expressing APEX on the ER membrane (facing the cytosol), we found that APEX-seq could clearly enrich secretory mRNAs on the ER membrane over cytosolic mRNAs, whereas APEX-RIP could not ( Figure 1D). Given the very close proximity between ER membrane-associated RNAs and cytosolic RNAs that are immediately adjacent to the ER, these results suggest that APEX-seq has a labeling radius of only a few nanometers.
Validation of APEX-seqEncouraged by the results above, we expanded APEX-seq to nine distinct subcellular locations ( Figure 1E). Correct localization of each APEX2 fusion protein (domain structures shown in Figure S2A) in HEK-293T cells was confirmed by imaging with organelle markers ( Figure 1F). We prepared two replicates for each location ( Figure S2B and S2C, correlation coefficient r between replicates ranged from 0.97 to 1) as well as negative controls in which H2O2 was omitted from the labeling reaction. After labeling for 1 minute and cell lysis, we enriched biotinylated RNAs and prepared APEX-seq libraries with polyA+ selection. Transcripts shorter than 100 nt were excluded during purification. We used DESeq2 (Love et al., 2014) for data analysis and used a qvalue (FDR-adjusted p-value) < 0.05 and a log2fold-change > 0.75 to select for significantly enriched transcripts.Consistent with the RT-qPCR results for the mitochondrial matrix shown in Figure 1C, our sequencing-based analysis showed that all 13 mRNAs and 2 rRNAs encoded by mtDNA are strongly enriched by MITO-APEX2 labeling (Figure 2A, S2D and S2E, mean enrichment > 11fold), whereas no mRNAs encoded by the nuclear genome are enriched. Figure 2B shows good correlation (r > 0.9) between technical replicates.To assess APEX-seq in an "open" subcellular region (not enclosed by a membrane), we focused again on the ER membrane (ERM). The ERM is particularly valuable for methodology validation because there are well-established "true positives" (mRNAs encoding secreted proteins that are known to be translated at the ERM) and well-established "false positive...